ARTICLE

https://doi.org/10.1038/s42003-020-01294-5 OPEN receptor 1 is a cellular receptor for human -

Brell Jennifer1,7, Verena Berg 1,7, Madhura Modak1, Alexander Puck1, Maria Seyerl-Jiresch1, Sarojinidevi Künig1, Gerhard J. Zlabinger 1, Peter Steinberger 1, Janet Chou2, Raif S. Geha2, Leopold Öhler3, Akihiro Yachie4, ✉ Hyeryun Choe 5, Markus Kraller6, Hannes Stockinger6 & Johannes Stöckl 1 1234567890():,;

Iron is essential for living cells. Uptake of iron-loaded transferrin by the transferrin receptor 1 (CD71, TFR) is a major but not sufficient mechanism and an alternative iron-loaded ligand for CD71 has been assumed. Here, we demonstrate that CD71 utilizes heme-albumin as cargo to transport iron into human cells. Binding and endocytosis of heme-albumin via CD71 was sufficient to promote proliferation of various cell types in the absence of transferrin. Growth and differentiation of cells induced by heme-albumin was dependent on heme-oxygenase 1 (HO-1) function and was accompanied with an increase of the intracellular labile iron pool (LIP). Import of heme-albumin via CD71 was further found to contribute to the efficacy of albumin-based drugs such as the chemotherapeutic Abraxane. Thus, heme-albumin/CD71 interaction is a novel route to transport nutrients or drugs into cells and adds to the emerging function of CD71 as a scavenger receptor.

1 Institute of Immunology, Center for Pathophysiology, Infectiology and Immunology, Medical University of Vienna, 1090 Vienna, Austria. 2 Division of Immunology, Boston Children´s Hospital, Boston, MA 02115, USA. 3 Department of Internal Medicine, St. Josef Hospital, 1130 Vienna, Austria. 4 Department of Pediatrics, School of Medicine, Institute of Medical, Pharmaceutical, and Health Sciences, Kanazawa University, Kanazawa, Japan. 5 Department of Immunology and Microbiology, The Scripps Research Institute, Florida, CA 92037, USA. 6 Institute of Hygiene and Applied Immunology, Center for Pathophysiology, Infectiology and Immunology, Medical University of Vienna, 1090 Vienna, Austria. 7These authors contributed equally: Jennifer Brell, ✉ Verena Berg. email: [email protected]

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 1 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

ron is an essential and multifunctional element for almost all Fig. 6). Labeling of plasma-derived HSA with heme strongly Iknown forms of life1–4. Numerous enzymes involved in DNA upregulated the proliferation-promoting capacity, whereas treat- replication, repair and translation rely on iron and therefore ment of HSA with different lipids did not enhance its proliferative iron is particularly indispensable for proliferating cells. For quality (Supplementary Fig. 8). HSA-heme is as potent as iron- example, DNA synthesis requires ribonucleotide reductase that loaded transferrin in promoting cell proliferation and there is no has an obligatory requirement for iron1. Transferrin is an abun- synergistic effect when both factors were applied in combination, dant plasma which binds iron and delivers it via binding suggesting that HSA-heme and transferrin may use the same and internalization through its receptor CD71 into cells. CD71 is transport route (Fig. 1e). absent or expressed at low levels on non-proliferating cells but is rapidly upregulated upon cellular activation and growth1,5,6. CD71 is the receptor for HSA-heme. To gain insight into a The transferrin/CD71 system is especially important for ery- potential role of CD71 in the binding and uptake of HSA-heme, throblasts and lymphocytes, T and B cells. Accordingly, defi- we used murine Bw-cells expressing human CD71 molecules and ciencies in this major iron-uptake route lead to anemia and analyzed the impact of HSA-heme on the proliferation of such compromised immune responses2,7. Surprisingly, several studies cells. Results presented in Fig. 2a demonstrate that recombinant have shown that defects of the receptor are more severe than – HSA-heme but not HSA alone facilitated the proliferation of defects or mutations affecting transferrin7 9. The reason for this murine BW-cells transfected with human CD71 but not of con- discrepancy is unclear but it is intriguing that CD71 might be also trol cells. Accordingly, murine splenocytes expressing mutated, used by other iron-carrying molecules. human CD71 receptors, which cannot internalize their cargo2,7, Albumin is the most abundant protein in our blood. It has the failed to proliferate in response to HSA-heme (Fig. 2b). Supple- unique ability to interact with numerous different substances mentation of culture medium with exogenous iron in form of such as lipids, drugs and also heme. Heme is iron-protoporphyrin ferric ammonium citrate (FAC) abrogated the need for HSA- IX. It is released from erythrocytes, mainly in the case of – heme in our protein-free culture condition (Fig. 2c). As demon- inflammatory or infectious diseases10 14. Since free heme is a strated in Fig. 2a, c, the proliferative responses of CD71- potential toxic substance for cells, it is immediately scavenged in transfected Bw cells, as well as of Jurkat T cells could be inhib- our body by plasma including , the pre- ited with mAbs against CD71. MAbs 5-528, 15-221 and VIP1 dominant binding partner, and albumin. It has been reported that recognize human CD71 expressed on murine Bw cells (Supple- human (HSA) has at least two binding sites for mentary Fig. 9). Blocking of CD71 with mAb VIP1 stopped heme. One site of HSA binds heme with high affinity and this proliferation induced by HSA-heme or of transferrin at the S/G2 binding is not affected by fatty acids15. The other site of HSA phase of the cell cycle (Supplementary Fig. 10). binds heme with low affinity and is used by various hydrophobic Using FITC-labeled proteins we could show that binding of molecules, including fatty acids16,17. The reported amounts of HSA-heme to Jurkat T cells was blocked with CD71 mAb VIP1 heme-albumin found regularly in our blood are in the range of − − and mAb 5-528, which recognize an overlapping epitope, but not 1×10 6 M to 1,5 × 10 6 M and this level is enhanced during − – with CD71 mAb 15-221 directed against a different epitope infection or inflammation to up to 5 × 10 5 M18 20. It is assumed (Fig. 3a, b and Supplementary Fig. 11). In addition, Jurkat T cells that albumin is binding about 45% of heme in our blood and that lacking CD71 on their cell surface could not bind HSA-heme up to 5% of albumin in humans is loaded with heme21. Given the (Fig. 3c). CD71 is not only a receptor for transferrin but also for abundance of HSA, a large amount of heme can be handled by Machupo virus, which binds via GP122,23. In order to investigate HSA, which might serve as an alternative, or perhaps additional if HSA-heme uses similar or different binding sites on CD71 mechanism to provide cells with iron. compared to the other ligands of the receptor, we performed binding studies and observed that transferrin reduces the binding Δ Results of HSA-heme to Jurkat T cells whereas MACV-GP1 -Ig (Machupo virus glycoprotein) did not affect binding of HSA- HSA-heme is able to promote cell proliferation. In order to test heme and was inhibited by mAb 15-221 but not VIP1 (Fig. 3d, whether HSA-loaded with hemin (HSA-heme) can be used by Supplementary Fig. 12). Thus, HSA-heme and transferrin have proliferating cells as an iron-source, we cultured Jurkat T cells overlapping binding sites on CD71 which are different from the with recombinant HSA or recombinant HSA-heme in serum-free epitope used by MACV-GP1Δ-Ig. Results show in Fig. 3e, f and protein-free medium. Results presented in Fig. 1a and Sup- demonstrate the internalization of HSA-heme in Jurkat T cells plementary Figs. 1 and 2 demonstrate that supplementing med- and a comparison with transferrin (Fig. 3e). Furthermore, ium with recombinant HSA-heme but not recombinant HSA proliferation induced by HSA-heme could be prevented with promotes proliferation of the cells. In contrast, HSA-loaded with inhibitors of clathrin-mediated endocytosis (MiTMAB, Dynasore, biliverdin (Fig. 1a) or with protoporphyrin IX (Supplementary Pitstop 2) (Fig. 3g). Fig. 3) is not sufficient for cell growth. The same proliferative response was observed with primary human T cells upon sti- mulation via CD3/CD28 or permanent cell lines such as ery- Heme oxygenase 1 is required for the utilization of HSA-heme. throblast TF-1 cells, monocytic THP-1 cells and epithelial Hela Cells use heme-oxygenase 1 to liberate iron from heme4,24.In cells (Fig. 1b, Supplementary Fig. 4). Addition of hemin alone in order to elucidate whether HO-1 is required that cells can use protein-free medium was not sufficient to induce proliferation in HSA-heme as an iron-source, we used lymphoblastoid YK01 cells cells (Supplementary Fig. 5). HSA from plasma but not bovine which are HO-1 deficient25. YK01 cells express high levels of serum albumin (BSA) was also capable of promoting cell pro- CD71 receptors like other lymphoblastoid B cell lines and pro- liferation, however, higher concentrations of the protein were liferation in the presence of iron-loaded transferrin is inhibited by needed compared to heme-loaded HSA (Fig. 1c). This effect was a CD71 mAb (Supplementary Figs. 13, 14). However, supple- observed with plasma-derived HSA from different sources, which menting the protein-free medium with HSA-heme is not able to carried heme but did not contain detectable amounts of trans- promote growth of YK01 cells in contrast to lymphoblastoid cells ferrin (Fig. 1d, Supplementary Figs. 6, 7). Treatment of HSA with OTHAKA with express intact HO-1 (Fig. 4a). Cell proliferation in charcoal largely removed heme from the protein and HSA lost its the presence of HSA-heme induces HO-1 expression (Fig. 4b). ability to promote cell proliferation (Fig. 1d, Supplementary The central role of HO-1 and the release of iron from HSA-heme

2 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

Fig. 1 HSA-heme is sufficient to promote proliferation under serum-free and protein-free culture conditions. a Effect of HSA-heme on the proliferation of Jurkat T cells. Cells were cultured in serum-free and protein-free medium supplemented with different proteins. Serum-free and protein-free RPMI medium was used as medium control (negative control) (number of experiments = 4). b Effect of HSA-heme on the proliferation of primary human peripheral blood T cells. T cells were activated via plate-bound mAbs CD3/CD28 and cultivated in medium supplemented with HSA or HSA-heme. Proliferation was measured after 4 days by analyzing [methyl-3H]thymidine incorporation (n = 3). c As in (a) the proliferation of Jurkat T cells was analyzed in presence of HSA or BSA (n = 2). d Jurkat T cell proliferation was analyzed in presence of HSA or charcoal treated HSA (n = 3). e Proliferation of Jurkat T cells in presence of HSA, HSA-heme and transferrin (n = 3). In addition, a counter-titration with transferrin and HSA-heme in combination was performed (n = 5). a–e Results are displayed standard error of the mean (SEM). P values were calculated by using one-way ANOVA, followed by Tukey’s multiple comparison test. *P < 0.05, **P > 0.01, and ***P < 0.001, ****P < 0.0001. was further examined by the use of an inhibitor. Results presented membrane permeable, but not with ethylenediamine tetraacetic in Fig. 4c demonstrate that proliferation of Jurkat T cells in the acid (EDTA), which primarily acts extracellularly (Fig. 5b)29–32. presence of HSA-heme but not fetal calf serum (FCS) is inhibited The increase of intracellular iron affected the expression of the by Tin Protoporphyrin, an inhibitor of HO-1. genes, which are regulated by the level of iron. Results presented In order to analyze whether the levels of the intracellular pool Fig. 5c demonstrate that TFR1 (CD71) and iron regulatory protein of LIP is altered in Jurkat T cells in the presence of HSA-heme, we 1 (IRP1) are downregulated in the presence of HSA-heme in used a calcein-based method26–28. We observed that HSA-heme Jurkat T cells, whereas ferritin is not significantly regulated, like enhanced the amounts of LIP like the treatment of Jurkat T cells we have observed in the case of adding iron in form of FAC. At with FAC (Fig. 5a). Moreover, HSA-heme-mediated cell pro- the protein level, HSA-heme induced a downregulation of TFR1 liferation can be blocked with iron-chelator 311, which is cell (CD71) expression but an upregulation of ferritin expression in

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 3 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

Fig. 2 CD71 is the receptor for HSA-heme. a Murine Bw cells lacking or expressing human CD71 receptors were cultured with 200 µg/ml HSA or HSA-heme in presence or absence of CD71 mAb VIP1 (10 µg/ml; n = 3). RPMI medium was used as medium control. b Proliferation of mouse spleenocytes, expressing mutated human CD71 (CD71mut), which cannot internalize their cargo. Spleenocytes were cultivated in RPMI medium supplemented with 10% FCS, HSA, or HSA-heme. In addition, cells were stimulated with PMA and Ionomycin and analyzed on day 5 via [methyl- 3H] thymidine incorporation (n = 4). c Effect of CD71 mAb VIP1 on the proliferation of Jurkat T cells. Cells were cultivated with 10% FCS, HSA, 10% FCS with FAC or HSA with FAC treated with or without VIP1 (n = 3). a, c Means ± SEM are shown. P values: *P < 0.05, **P > 0.01, and ***P < 0.001 (two-tailed unpaired t-test). b Means ± SEM are shown. ****P < 0.0001 (one-way ANOVA followed by Tukey’s multiple comparison test).

via CD71 is not only an inert nutrient and iron-source for cells but may influence cellular functions via down-stream degradation products of heme. Using a Jurkat reporter cell line, termed triple parameter reporter (TPR), we next analyzed whether HSA-heme influences three prominent signaling pathways (NF-κB, AP1, NFAT)35. Results presented in Fig. 6a demonstrate that Jurkat cells activated via CD3/CD28 in the presence or absence of plasma-derived HSA or HSA-heme turn-on NF-kB, AP1 and NFAT like the cells do in the presence of transferrin. HO-1 generated degradation products of heme are also known to affect the differentiation and function of the myeloid cell linage36,37. Myeloid cells, such as dendritic cells, upregulate CD71 expression upon differentiation and activation5,6. We observed that HSA- heme regulates the functional differentiation of monocyte-derived DCs. DCs generated in the presence of HSA-heme expressed lower levels of the typical marker CD1a compared to DCs gen- erated in the presence of transferrin or FCS (Fig. 7a). The induction of production of several cytokines analyzed in this study (IL-12, IL-10, IL-1, and TNF) induced by LPS stimulation was reduced in HSA-heme cultured DCs, in particular IL-10 and IL-12p40 (Fig. 7b).

The toxic effect of Abraxane is part mediated via CD71. Albumin is a multifunctional protein and a potential carrier and scavenger of many different compounds such as lipids and drugs38–40. Human plasma-derived albumin is therefore applied as carrier and stabilizer in many formulations of medicines. Abraxane is an HSA-based drug that consists of paclitaxel, a chemotherapeutic agent, which is packed into HSA nano- particles40,41. Abraxane contains heme-moieties (Fig. 8a). Based on our observations we asked whether the CD71 route, which is also used by large molecular complexes such as viruses, may be involved in the uptake of Abraxane. The drug is toxic for Jurkat T cells down to 5 ng/ml (Fig. 8b). The toxic effect of Abraxane is caused by preventing microtubule depolymerization42. Reactivity of anti-beta-tubulin is enhanced in Abraxane-treated cells and this effect is inhibited by CD71 mAb 5-52843 (Fig. 8c). Moreover, blocking of CD71 with mAb reduces the number of dying cells by approximately 25% (Fig. 8d). Collectively, these results reveal a novel role of HSA-heme and CD71 in delivering iron to cells and suggest an impact of the scavenger function of Jurkat T cells (Fig. 5d). Thus, HSA-heme can provide cells with CD71 in the mechanism of HSA-based drug applications. iron from heme catabolism involving HO-1. Discussion Impact of HSA-heme on cellular signaling and differentiation. Albumin is somehow a forgotten stepchild, although it is the most HO-1 degrades heme into biliverdin which is then further abundant intravascular protein and a substantial extravascular degraded into bilirubin and other down-stream products protein. HSA is responsible for 80% of the colloidal osmotic including CO24,33,34. Hence, we asked if HSA-heme transported pressure of blood. It is surprising that in spite of this impressive

4 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

Fig. 3 HSA-heme uptake is mediated by endocytosis. a Effect of CD71 mAbs on HSA binding to CD71. Binding of FITC-labeled HSA (100 µg/ml) to CD71 was analyzed by flow cytometry. Jurkat T cells were pre-treated with or without CD71 mAb VIP1, 5-528, or 15-221. b Bar graph shows the mean fluorescence intensity (MFI) of HSA binding to CD71. Data shown are representative of four independent experiments. c Analysis of the binding of HSA- FITC to CD71 in Jurkat T cells and Jurkat T cells, which lack CD71 surface expression due to mAb-induced down-modulation. Fluorescence was measured by flow cytometry. Data presented is representative of three independent experiments. d FACS analysis of the binding and blocking of HSA and GP1ΔIg protein to CD71. Jurkat T cells were pre-treated with transferrin and incubated with FITC-labeled HSA (200 µg/ml) or GP1ΔIg (6 µg/ml) protein, followed by a second step staining with a secondary antibody. Data shown is representative for four independent experiments. e FACS analysis of the internalization of HSA-FITC and Tfn-FITC at 37 °C (n = 4). f HSA internalization was visualized by Epifluorescence microscopy on live cells. The uptake (37 °C) and binding (0 °C) of AF488-labeled HSA was assessed. Differential interference contrast (DIC) and fluorescence micrographs are shown. Results represent one of three independent experiments. Scale bar: 5 µm. g Jurkat T cells were cultivated in HSA-heme at a concentration of 200 µg/ml. In comparison to the control (Mock), cells were additionally treated with inhibitors MiTMAB (n = 3), Dynasore (n = 3), or Pitstop 2 (n = 5) at different concentrations. b, g Data show mean ± SEM. P values were calculated by using one-way ANOVA, followed by Tukey’s multiple comparison test. *P < 0.05, **P > 0.01, and ***P < 0.001. plethora of functions, people suffering from hypo-albuminemia Accordingly, although hemopexin is considered as the primary or from mutated HSA, have seemingly no serious health problem. scavenger of free heme molecules, the high amounts of HSA These observations also demonstrate that HSA is obviously not of molecules guarantee an efficient buffer system of the toxic effects a single importance for our body and may rather serve as a of free heme on cells and HSA binds about 30% more free heme molecular back-up system for several molecular functions. than hemopexin11,21.

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 5 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

Fig. 4 Utilization of HSA-heme by proliferating cells requires heme oxygenase 1 (HO-1). a Proliferation of Epstein-Barr-Virus (EBV)- immortalized B cells, a wildtype (OTHAKA) and a cell line with a defect heme oxygenase 1 enzyme (YK01) in presence of HSA or HSA-heme (n = 3). RPMI medium without protein or serum was used as medium control. b HO-1 gene expression in Jurkat T cells. Levels of the indicated HO-1 mRNAs in the presence of HSA or HSA-heme (200 µg/ml) were quantified via qPCR after 4 h, 24 h, and 48 h. For each time point three independent experiments are shown. c Jurkat T cells were incubated in serum-free and protein-free medium supplemented with 10% FCS (Mock) or HSA-heme at a concentration of 200 µg/ml. Cells were treated with Tin Protoporphyrin (HO-1 inhibitor) in three different concentrations (n = 3). a–c Means ± SEM are shown. *P < 0.05, ***P < 0.001, (one-way ANOVA followed by Tukey’s multiple comparison test).

uptake of HSA-heme can be utilized by different cell types to gain access to an alternative source of iron sufficient to promote their proliferation. We have shown this effect for primary human T cells, as well as for a panel of immortalized cell lines. Yet, it is intriguing to speculate that different cell types may differ in their capacity to catabolize and utilize HSA-heme as nutrient. Intact HO-1 function is important for cells to utilize iron after uptake of HSA-heme (Fig. 4a–c). Yet, degradation of heme by HO-1 is also an important cellular differentiation signal. We have observed in this study that HSA-heme is able to modulate the differentiation of monocyte-derived DC and to inhibit their cytokine production repertoire (Fig. 7b). So, it needs to be determined in future studies whether provision of iron via HSA-heme/CD71 mediated endo- cytosis in a general phenomenon in our cells and how this pathway may influence the functional behavior of cells. CD71 is considered today as a promiscuous cell entry carrier. Its primary function is the import of iron bound to transferrin but several other ligands including ferritin, arenaviruses or malaria parasite use CD71 to enter cells. Our study demonstrates that HSA-heme is another intrinsic complex that utilizes this special cell entrance. The dissociation constant (Kd) value of HSA-heme binding is 7,52 × 10−7 M, which is lower range of what has been reported for transferrin. The Kd for bound diferric transferrin ranges from 10−7 Mto10−9 M at physiologic pH, depending on the species and tissue. Based on our binding studies, we conclude that the binding site of HSA-heme is in proximity to the trans- ferrin binding site, which is located on the lateral part of the receptor ectodomain. This binding site is distinct from the fer- ritin/pathogen contact region of CD71, which is located on the apical region of CD71 ectodomain. Interestingly, the HSA-heme/ CD71 route seems to be specific for human albumin and human CD71. Our results of this study demonstrate that plasma-derived BSA, which is also loaded with heme although not so frequent as human albumin44,45, did not promote cell proliferation in our test system. Moreover, the data of our study obtained with murine cells revealed that HSA-heme can obviously not be used as an iron source (Fig. 2a, b). Thus, it seems as if the albumin/CD71 interaction could be species specific. Iron is one of the central chemicals in our body and we demonstrate here a novel route of iron-uptake via CD71. How- ever, we have to envisage that CD71 molecules exert also other, non-canonical functions. Mouse genetic studies have implicated Our body and cells have to remove HSA-heme from the sys- iron-independent roles of CD71 in the early development of the tem. Several receptors for HSA have been reported for various cell nervous system46, maintenance of intestinal cell homeostasis47 types13. Here we demonstrate that CD71 is an important and and development of lymphocytes48. In addition, CD71 has sig- specific cellular receptor for HSA-heme. The CD71/transferrin naling functions, including the activation of NF-κB and regula- system is famous for iron-transport into proliferating cells and is tion of JNK35. It will be worth testing whether HSA-heme, an a prototypic endocytic receptor, which is recycling to the cell alternative ligand of CD71, may be involved in these non- surface. The results of our study demonstrate that binding and canonical roles of CD71. Interestingly, the function of CD71 can

6 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

Fig. 5 Iron from HSA-heme is used for cell proliferation. a Impact of HSA-heme on intracellular levels of the labile iron pool (LIP). Jurkat T cells were incubated for 2 h with HSA-heme or FAC. Cells were loaded with Calcein-AM, washed and incubated with a combination of iron chelators: 311 (Fe3+ chelator) and BIP (Fe2+ chelator). Data show mean fluorescence between chelator-treated and untreated cells (Δ MFI). b Jurkat T cells were incubated in medium supplemented with 10% FCS (Mock) or HSA-heme at a concentration of 200 µg/ml. In addition, cells were treated with iron chelator 311 (n = 4) or EDTA (n = 3) at different concentrations. c TFR1, IRP1 and ferritin mRNA expression under different conditions. Jurkat T cells were incubated with 10% FCS, HSA-heme (200 µg/ml) or 10% FCS with FAC (25 µg/ml) for 6 h. Expression of mRNAs were quantified via qPCR and mRNAs were normalized to β2 m mRNA. Results are from three (TFR1, IRP1) or four (ferritin) independent experiments. d Expression of human TFR (CD71) and ferritin at the protein level. As in (c) Jurkat T cells were incubated with HSA-heme or FAC and protein expression was detected after 24 h (TFR) or 48 h (ferritin) by intracellular staining with CD71 mAb and anti-ferritin antibody and analyzed by flow cytometry. Data show mean values of MFI.a–d Means ± SEM are shown. **P > 0.01, ***P < 0.001, ****P < 0.0001 (two-tailed unpaired t test).

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 7 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

Fig. 6 Impact of HSA-heme on cell signaling. a A multi-channel reporter cell line from Jurkat T cells expressing reporter genes under control of NF-κB, NFAT, and AP-1 promoter elements were used to study the impact of HSA-heme on prominent signaling pathways. Bar graph shows the expression of the transcription factors NF-κB, NFAT, and AP1 in presence of 10% FSC, HSA, HSA-heme or transferrin analyzed by flow cytometry. Bar graph show resting cells and cells which were stimulated with a combination of CD3 and CD28 (5 µg/ml each). Graph show means ± SEM of MFI (n = 4). Only significant differences are indicated (one-way ANOVA followed by Tukey’s multiple comparison test). be modulated by stearoylation, which regulates the morphology Machupo virus, cause hemorrhagic fever, the elevated levels of of mitochondria49 or by c-Abl kinase, which controls the endo- HSA-heme in such a situation may contribute to protect the cytic fate of CD7150. This finding indicates that the interaction CD71 entry site. partners of CD71 may regulate its functional repertoire. HSA- Albumin is an important drug carrier is gaining increasing heme has peroxidase-activity51. Peroxidase-activity is a well- importance in the target delivery of cancer therapy, particularly as known anti-microbial tool of our innate immune system52–54. a result of the market approval of the paclitaxel-loaded albumin Binding of HSA-heme to CD71 may thus decorate this important nanoparticle, Abraxane40,41. CD71 is an entry gate for large portal of viruses and malaria parasites with peroxidase activity. It molecular particles such as viruses55 The results of this study that is intriguing to speculate that since some of these viruses, such as approx. 25% of the toxic efficacy of Abraxane is mediated via the

8 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

herein described uptake mechanism of HSA-heme via CD71. Accordingly, we could demonstrate that CD71 mAb 5-528 reduces the increased beta-tubulin staining in cells upon Abraxane-treatment. The binding sites of mAb 5-528 and VIP1 are overlapping (Supplementary Figs. 9 and 10). Both mAbs inhibit binding of HSA-heme to CD71, where 5-528 is less effi- cient (Fig. 3a, b). However, mAb VIP1 inhibits cell proliferation (Fig. 2), whereas mAb 5-528 does not. Thus, the inhibitory mAb VIP1 is not suitable to revert the cell toxic effect of Abraxane, whereas 5-528 is able to partly revert the Abraxane killing because the mAb is not affecting the proliferation and viability of the cells per se. Taken together, the HSA-heme/CD71 route described in this study offers new insights of iron-uptake of cells, the design of future HSA based drugs and the usage of HSA as plasma- compensation.

Methods Cell lines and culture. All the cell lines used in this study were tested for mycoplasma contaminations using a method described before56. Jurkat T cells (TIB-152), TF-1 cells (CRL-2003), THP1 cells (TIB-202), HeLa cells (CCL-2), Bw5147 cells (TIB-47) were obtained from ATCC (American Type Culture Col- lection, Virginia, US). A multi-channel reporter cell line from Jurkat T cells expressing reporter genes under control of NF-κB, NFAT and AP-1 promoter elements and murine Bw cells transfected with human CD71 were generated at our institute as described37,57. The EBV-transformed lymphoblastoid B cell lines YK01 (HO-1 deficient) and OTHAKA were generated as described before25,58. Jurkat T cells which do not express CD71 on the cell surface (CD71 deficient Jurkat T cells) were generated due to down-modulation of cell surface molecules by mAb VIP1 treatment and subsequent culturing in the presence of exogenous iron in form of FAC. All cell lines, were cultured in RPMI 1640 medium (Invitrogen, Paisley, United Kingdom), supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin and 10% FCS purchased from HyClone fi Laboratories Inc. (Utah, US), in a humidi ed atmosphere (5% CO2) at 37 °C. In case of TF-1 cells 2 ng/ml GM-CSF was additionally added to RPMI 1640 medium. All of the above-mentioned cell lines are immortalized or tumor-derived cell lines.

Antibodies and reagents. The following murine monoclonal antibodies (mAb) were raised in our laboratory: negative control mAb VIAP (against calf intestine alkaline phosphatase), 5-528 (CD71), VIP1 (CD71)59,60, 15-221 (CD71), 13-344 (transferrin), mAb VIT6b (CD1a-PE), 5-272 (CD274-PE), MEM 18 (CD14-PE), and 7-239/44/0 (CD169-PE). APC-conjugated donkey anti-human IgG, goat anti- human IgG and anti-mouse conjugated alkaline phosphatase antibodies were purchased from Jackson-Immunoresearch Laboratories Inc. (Newmarket, UK). mAb OKT3 (CD3) was obtained from Janssen-Cilag (Vienna, AT). Human granulocyte–macrophage colony-stimulating factor (GM-CSF) and IL-4 were kindly provided by Novo Nordisk A/S (Bagsværd, DNK). MAb B7-2 (CD86-PE), mAb 10F3 (CD28), Oregon Green 488-conjugated goat anti-mouse IgG antibody and UltraPure EDTA was obtained from Invitrogen, UK. Native HSA-FITC, MiTMAB and Deferiprone were purchased from Abcam PLC (Cambridge, UK). Anti-human beta-tubulin antibody (9F3-AF488) was acquired from Cell Signaling Technology Inc. (Frankfurt, DEU). Anti-human ferritin antibody (heavy chain, ferritin-AF647) was purchased from Santa Cruz (Dallas, US) and anti-human TFR antibody (TFR-AF647) from BD (Vienna, Austria). The following products were purchased from Sigma-Aldrich (St. Louis, US): transferrin (Tfn-FITC), phorbol myristate acetate (PMA) and ionomycin, lipopolysaccharide (LPS) from E. coli 0127:B8, FAC, holo-transferrin, linoleic acid, oleic acid, hemin (porcine), bili- Fig. 7 Impact of HSA-heme on the differentiation of monocytes into DCs. verdin-hydrochlorid, AS8351 (311), Protoporphyrin IX, Dynasore hydrate, Pitstop a Phenotype of DCs generated in the presence of HSA-heme. Monocytes 2, 2,2 Bipyridyl (BIP), propidium iodid and calcein-acetoxymethyl ester (Calcein- AM) was obtained from Biozyme Scientific GmbH (Vienna, Austria). Tin Proto- were differentiated into DCs in presence of 10% FCS, 1% HSA or 1% HSA- porphyrin IX was from Bio-techne Ltd (Abingdon, UK). GP1Δ-Ig (Machupo virus heme and analyzed after 7 days for expression of typical cell surface glycoprotein) and the control protein SNIT were generated as recently described22. markers (filled histograms) compared to the isotype control (dotted Abraxane was obtained from Celgene GmbH (Summit, US), FIX and PERM® from histograms) via flow cytometry. Data shown are representative of four Nordic-MUbio (Susteren, NLD) and [methyl-3H]-thymidine from Perkin Elmer/ independent experiments with four different donors. b HSA-heme New England Corporation (Wellesley, MA). modulates cytokine production of DC upon stimulation with LPS (mature DCs). Supernatants were collected from DCs and mature DCs and analyzed Serum-free and protein-free medium. Cells were maintained in RPMI 1640 n = *P medium, supplemented with 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ on day 7 ( 3). Means ± SEM are shown. < 0.05 (one-way ANOVA ml streptomycin without FCS. The protein-free medium was further supplemented followed by Tukey’s multiple comparison test). with different HSA proteins, as mentioned in the text.

Albumin proteins. In this study we have used two proteins (HSA) which were plasma-derived from human blood: HSA (Albiomin) from Biotest (Dreieich, DE), which is has clinical grade, and HSA from Sigma-Aldrich (St. Louis, US). Fatty acid free HSA (dHSA) was purchased from Sigma-Aldrich,

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 9 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

Fig. 8 Abraxane induced cell death is executed in part via CD71. a Bar graph illustrates the concentration of heme (µM), which was bound to HSA and HSA bound Paclitaxel particles (Abraxane) at different concentrations via a Heme Assay Kit (n = 3). b Toxic effect of Abraxane on Jurkat T cells. Cells were incubated with Abraxane for 6 days (n = 3). RPMI medium without serum was used as medium control. c Analysis of anti-beta-tubulin antibody binding to Jurkat T cells by flow cytometry. Data show mean values of MFI of anti-beta-tubulin-antibody binding to beta-tubulin in presence or absence of Abraxane (100 ng/ml), with or without mAb 5-528 (CD71). Data shown is representative of three independently performed experiments. d Effect of CD71 mAbs on Abraxane-treated cells. Jurkat T cells were incubated with Abraxane (100 ng/ml) in combination of CD71 mAbs (10 µg/ml), VIP1, 5-528, 15-221, or 13-344 (control). After 72 h, cell viability was analyzed by PI staining and flow cytometry (n = 3). b, d Error bars represent standard deviation (SD). *P < 0.05 and **P < 0.01 (one-way ANOVA followed by Tukey’s multiple comparison test). which was generated from HSA (Sigma-Aldrich) due to charcoal treatment. molecules, 251 µg/ml iron) and transferrin (1,5 × 1015 protein molecules, 0.3 µg/ml Recombinant HSA expressed in S. cerevisiae (rHSA) or in Oryza sativa (OSrHSA) iron) carry 2,7 × 1018 and 3 × 1015 iron molecules, respectively. was acquired from Sigma-Aldrich. BSA was purchased from GE Healthcare HSA was labeled with Alexa Fluor 488 (NHS; HSA-AF488) dye according to the (Pasching, AT). The endotoxin levels in all recombinant probes was <1EU/mg. manufacturer´s instructions (Thermo Fisher Scientific Inc., Waltham, US). After labeling a purification run using a S200 column (Superdex 200, 10/300 GL, GE Healthcare Life Sciences) on an ÄKTA purifier system (GE Healthcare Life Cell isolation and stimulation . Buffy coats from healthy donors were purchased Sciences, Pittsburgh, US) was performed. Protein was concentrated and set to a either from the Austrian Red Cross or University Clinic for Blood Group Serology concentration of 12 mg/ml. and Transfusion Medicine, Medical University of Vienna (both, Vienna, Austria). Peripheral blood mononuclear cells (PBMC) were isolated from heparinized buffy coats via standard density gradient centrifugation using Ficoll‐Paque™ Plus (GE Heme-assay kit. To quantify the amount of hemin bound to different HSA a Healthcare, Chalfont St Giles, US). T (CD3+) cells and monocytes were purified Heme Assay Kit protocol was used. Heme Assay Kit (MAK316) was purchased from PBMCs using the MACS system (Miltenyi Biotec, Bergisch Gladbach, DEU). from Sigma-Aldrich (St. Louis, US). Hemin level was determined using manu- T cells (total CD3+ cells) were obtained via depletion of CD11b, CD14, CD16, facturer’s protocol and calibration standards provided. In brief, HSA and Abraxane CD19, CD33, and MHC class II-positive selection using biotinylated CD14 mAb. were diluted at different concentrations and applied to 96-well plates. The reaction Dendritic cells (DCs) were generated by culturing purified blood monocytes for mix (supplied with the kit) was added, and the plates were incubated for 5 min at 6 days with a combination of GM-CSF (50 ng/ml) and IL-4 (35 ng/ml). On day 6 of room temperature in the dark. Changes in absorbance at 405 nm were measured differentiation, DCs were stimulated using 1 µg/ml LPS for 24 h. Murine splee- using a 96-well plate reader (Bio-Rad Laboratories, Hercules, CA). nocytes (WT) and TFRC knockin spleenocytes (KI) from mice expressing mutated human transferrin receptor were generated as described previously7. Culture and activation of T cells. MAXISORP Nunc-Immuno plates (Thermo HSA, hemin, biliverdin, and protoporphyrin preparation and conjugation. Scientific, Waltham, MA) were coated overnight at 4 °C with CD3 mAb (5 µg/ml) Hemin, biliverdin, and protoporphyrin were dissolved in 1 M NaOH (3 × 10−7 M) in combination with CD28 (5 µg/ml) mAb. The plates were then washed to remove and incubated with 10 mg/ml HSA for 1 h at room temperature. The protein was unbound mAbs and purified T cells (2 × 105/well) were then added to the dialyzed for 48 h against 1× PBS, changing the buffer after 24 h, to remove non- respective wells. T cell proliferation was assessed via cellular incorporation of bound hemin. HSA-heme, HSA-biliverdin and HSA-protoporphyrin were then [methyl-3H]-thymidine. Cells were labeled with 0.05 mCi/well of [methyl-3H]- sterile filtered. The loading of HSA with hemin was analyzed after filtration by thymidine on day 3 and cultured for another 18 h prior to harvesting. Detection measuring the absorbance spectra. The ratio of HSA: heme was 1: 1,5 ± 0,36 was performed on a microplate counter (MicroBeta2; Perkin). All assays were (n = 8). At a protein concentration of 200 µg/ml, HSA-heme (1,8 × 1018 protein performed in triplicates and readings are displayed as counts per minute.

10 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

Flow cytometry. For cell surface staining, 5 × 105 cells per staining were incubated RNA isolation and qPCR. Total cellular RNA was isolated using peqGOLD TriFast with either unconjugated mAb, PE-conjugated or FITC-conjugated mAb for reagent (peqLab, Erlangen, DEU). For isolation, 5 × 105 cells/ml were lysed in 30 min at 0 °C. For unconjugated mAbs or proteins (GP1ΔIg), secondary Oregon 500 µl of TriFast reagent and isolated according to the manufacturer’s protocol. Green 488 (IgG) or APC-conjugated donkey anti-human IgG was added to samples 1 µg of total RNA per sample was reverse transcribed using H-Minus-Reverse fi as second step reagents. In some experiments, cells were pretreated with different Transcriptase (Thermo Fisher Scienti c Inc., Waltham, US) and oligo-dT18 pri- antibodies or proteins like CD71 mAbs (5 µg/ml), transferrin (50 µg/ml), Abraxane mers. Quantitative real-time PCR was performed with a CFX96 Real-Time PCR (100 ng/ml), HSA or HSA-heme (both 200 µg/ml). KD value for HSA-heme was Detection System (Bio-Rad, Hercules, CA) using SYBR Green qPCR master mix calculated as described61. Flow cytometry analyses were performed using FACS- (Quanta Biosciences, Gaithersburg, MD) for detection. Detection was performed Calibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ). according to the manufacturer’s protocol. cDNA was amplified using a standard Intracellular stainings were performed by fixing cells for 20 min at room program (10 min at 95 °C, 40 cycles of 15 s at 95 °C/15 s at 60 °C/45 s at 72 °C). temperature using FIX solution followed by permeabilization in the presence of CD3E or β2m was used as an endogenous reference gene. Primers specific for anti-beta−tubulin-AF488 (1 µg/ml) using PERM solution for 20 min at room HO-1, TFR1, TF, IRP1 and ferritin were designed using Primer 3 Plus software and temperature. These cells were pretreated with Abraxane (100 ng/ml) for 2 h at were custom synthesized by Sigma-Aldrich. Sequences were as follows: HO-1 fwd 37 °C with or without 10 µg/ml mAb 5-528. 1% BSA/PBS was used as staining 5′-AAGATTGCCCAGAAAGCCCTGGAC-3′ and HO-1 rev 5′-AACTGTCGCC buffer. DNA staining was performed with propidium iodide. ACCAGAAAGCTGAG-3′; TFR1 fwd 5′- ACTTCTTCCGTGCTACTTCCAG-3′ and TFR1 rev 5′-ACTCCACTCTCATGACACGATC-3; TF fwd 5′-ATGCTGCC ACCTCTAGAATGTC-3′ and TF rev 5′-TTTGGGCTTTTGGACTCAGC-3′; IRP1 Protein concentration . Some intracellular stainings were performed with cyto- fwd 5′-TCCCTGGTGAGAATGCAGATG-3′ and IRP1 rev 5′- TCTTGCCAGTA plasmatic antibodies. Jurkat T cells were cultivated for 24 or 48 h with 10% FCS TCCAGCTTGAC-3; ferritin fwd 5′-CAGAACTACCACCAGGACTCAG-3′ and medium with or without HSA-heme or FAC (25 µg/ml). After 24 h expression of ferritin rev 5′- GTCATCACAGTCTGGTTTCTTG-3. Data analysis was performed human TFR1 and after 48 h expression of human ferritin was assessed by using using CFX Manager Software (Bio-Rad). Alexa 647 conjugated anti-human TFR1 and Alexa 647 anti-human ferritin Jurkat T cells do not produce transferrin. After 6 h and 24 h of culture in the antibody. presence of either FCS, HSA, HSA-heme or FAC, expression of transferrin mRNA was analyzed by qPCR and not detectable in the Jurkat T cells. Internalization. For endocytosis cells (5 × 105) were incubated with different concentrations of HSA-FITC or Tfn-FITC at 0 °C or 37 °C for 1 h and were ana- Cell cycle analysis. Jurkat T cells were harvested on day 4 of culture and were lyzed by FACS. For microscopy, Jurkat T cells were exposed to 1 mg/ml HSA- fixed in 70% ethanol at 4 °C for 30 min. After washing, cells were stained with AF488 for 1 h at 0 °C or 37 °C for internalization. 1% BSA/PBS was used as 50 μg/ml propidium iodide in the presence of 100 μg/ml RNAse A for 45 min at fl washing buffer. Epi uorescence microscopy was performed with an Eclipse Ti-E room temperature. DNA content was then measured on a BD FACSCalibur flow fi (Nikon) inverted microscope system with a high NA objective (×100 magni cation, cytometer (BD Biosciences, San Jose, CA). NA = 1.49, Nikon SR APO TIRF). The sample was illuminated by using a 488 nm diode laser (iBeam smart Toptica). Appropriate filters to visualize GFP were used. For recording the emission light we used a backside-illuminated EM-CCD camera Multi-channel reporter cell line assay. To analyze the activation of downstream (Andor iXon Ultra 897). Metamorph (Molecular Devices) was used to control signaling pathways, a multi-channel reporter cell line Jurkat E6 expressing reporter κ hardware elements. genes under control of NF- B, NFAT, and AP-1 promoter elements was used. The promotor elements drive the expression of fluorescent proteins (eCFP, eGFP, mCherry). The reporter cell line was generated by introducing constructs encoding Determination of the intracellular LIP level. Jurkat T cells were incubated for 2 h NF-κB-eCFP, NFAT-eGFP, and AP-1-mCherry into Jurkat E6 cells. A cell clone with 10% FCS medium, HSA-heme (200 µg/ml) or FAC (25 µg/ml). At the used that was negative for fluorescent proteins in an unstimulated state and strongly concentrations, the amount of iron was 251 µg/ml and 5 µg/ml for HSA-heme and upregulated eCFP, eGFP and mCherry expression upon PMA/ionomycin treat- FAC, respectively. Cells were collected and washed twice with 1% BSA/PBS ment was selected for further use. The reporter cells were cultivated in serum-free (washing buffer) and incubated at a density of 2 × 105 cells for 15 min at 37 °C in medium supplemented with HSA, HSA-heme or transferrin and were activated via the dark with 30 nM calcein-AM. Then, the cells were washed twice and treated + + plate bound CD3 and CD28 mAb at a concentration of 5 µg/ml. To assess the with a combination of 10 µM 311 (Fe3 chelator) and 250 µM BIP (Fe2 chelator) activation of the respective transcription factors, cells were harvested after 4 h, 24 h, for 1 h at RT. Afterwards, cells were analyzed by flow cytometry. The LIP level was and 48 h and expression of the receptor genes eCFP, eGFP, and mCherry were calculated as the difference in mean fluorescence intensity of the cells treated with measured by flow cytometry using a LSRFortessa (flow cytometer, Becton or without 311/BIP (Δ MFI). Dickinson).

ELISA assay. Transferrin was detected and measured by ELISA. Flat bottomed, 96 Cytokine measurements. Supernatants from monocyte-derived DC were collected well ELISA plates (Corning Life Sciences, Tewskbury, MA) were coated with 2% after 24 h of LPS stimulation or from unstimulated cells. Cytokines were deter- blood derived HSA or 100 µg/ml transferrin overnight at 4 °C. The plate was mined using manufacturer’s protocol. In brief, supernatants were prepared, the washed twice with a PBS-Tween 0.05% solution. Blocking was performed by reaction mix (Multiplex Assay, Merck Millipore) was added and cytokines adding 200 µl of a 2% PBS-BSA solution to each well and incubating again over- including IL-10, IL-12p70, IL-6, and TNF-α were measured by using Luminex 100 night at 4 °C. After washing with PBS-Tween, 50 µl of mAb 13-344, directed System (R&D Systems Inc.). All measurements were performed in triplicates. against transferrin, was added and incubated for 2 h at 4 °C in the dark. Transferrin was used as positive control. The plate was washed then 3 times with PBS-Tween. Statistics and reproducibility. Statistical analysis was performed using GraphPad Afterwards, the 13-344 was detected with an anti-mouse antibody labeled with Prism software (La Jolla, CA). One-way analysis of variance (ANOVA) followed by alkaline phosphatase (1:2000) and incubated for 1 h at room temperature. Tukey’s multiple comparison test or unpaired, two-tailed Student’s t-test, followed Unbound antibodies were removed by washing 3 times. Substrate buffer containing by Holm-Šídák test for multiple comparisons, was performed. The P-values <0.05 diethanolamine was added to analyze the bound antibodies colorimetrically using were considered significant and are represented as *P < 0.05,**P < 0.01, ***P < an ELISA reader (Bio-Rad Laboratories, Hercules, CA). 1% NaOH was used as stop 0.001, ****P < 0.0001. For all statistical analysis data from at least two biological solution. repeats performed on separate days was used. The exact number of replicates are presented in individual figure legends. Any differences in statistical significance are Proliferation assay. Jurkat T cells were cultured at 3 × 104 – 1 × 105/well in 24- indicated. well plates (Costar, Sigma Aldrich) using RPMI 1640 medium supplemented with various proteins (10% FCS, HSA, BSA, HSA-heme, HSA-biliverdin, HSA-proto- Reporting summary. Further information on research design is available in the Nature porphyrin, human transferrin, heme, biliverdin) at different concentrations instead Research Reporting Summary linked to this article. of 10% FCS or 10% FCS with FAC. In some experiments anti-human CD71 mAbs (10 μg/ml) were added. Fatty acids, linoleic acid and/or oleic acid were used at a final concentration of 100 mM. Abraxane, MiTMAB, Dynasore, Pitstop 2, EDTA, Data availability 311 or Tin Protoporphyrin were applied at different concentrations. Proliferation All source data underlying the graphs and charts in the main figures are available in the was monitored with light microscopy. Cells were harvested at day 6 and the cell Supplementary Data section. The data that support the findings of this study are available viability was measured in triplicates with CASY cell counter (Roche Innovatis AG, from the corresponding author (J.S.) upon reasonable request. Bielefeld, Germany). In other experiments cells were assessed via cellular incor- poration of [methyl-3H]-thymidine. Cells were labeled (0.05 mCi/well) of [methyl- 3H]-thymidine on day 3 and cultured for another 18 h prior to harvesting. Received: 13 May 2019; Accepted: 15 September 2020; Detection was performed on a microplate scintillation counter. All assays were performed in triplicates.

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 11 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5

References 32. Richardson, D., Ponka, P. & Baker, E. The effect of the iron(III) chelator, 1. Richardson, D. R. & Ponka, P. The molecular mechanisms of the metabolism desferrioxamine, on iron and transferrin uptake by the human malignant and transport of iron in normal and neoplastic cells. Biochim. Biophys. Acta melanoma cell. Cancer Res. 54, 685–689 (1994). 1331,1–40 (1997). 33. Lee, P. J., Alam, J., Wiegand, G. W. & Choi, A. M. Overexpression of heme 2. Trenor, C. C. 3rd, Campagna, D. R., Sellers, V. M., Andrews, N. C. & Fleming, oxygenase-1 in human pulmonary epithelial cells results in cell growth arrest M. D. The molecular defect in hypotransferrinemic mice. Blood 96, 1113–1118 and increased resistance to hyperoxia. Proc. Natl Acad. Sci. USA 93, (2000). 10393–10398 (1996). 3. Hentze, M. W., Muckenthaler, M. U. & Andrews, N. C. Balancing acts: 34. Chauveau, C. et al. Heme oxygenase-1 expression inhibits dendritic cell molecular control of mammalian iron metabolism. Cell 117, 285–297 (2004). maturation and proinflammatory function but conserves IL-10 expression. 4. Soares, M. P. & Weiss, G. The Iron age of host-microbe interactions. EMBO Blood 106, 1694–1702 (2005). Rep. 16, 1482–1500 (2015). 35. Kenneth, N. S., Mudie, S., Naron, S. & Rocha, S. TfR1 interacts with the IKK 5. Ponka, P. & Lok, C. N. The transferrin receptor: role in health and disease. Int. complex and is involved in IKK-NF-kappaB signalling. Biochem J. 449, J. Biochem. Cell Biol. 31, 1111–1137 (1999). 275–284 (2013). 6. Daniels, T. R., Delgado, T., Rodriguez, J. A., Helguera, G. & Penichet, M. L. 36. Wegiel, B. et al. Heme oxygenase-1 derived carbon monoxide permits The transferrin receptor part I: biology and targeting with cytotoxic antibodies maturation of myeloid cells. Cell Death Dis. 5, e1139 (2014). for the treatment of cancer. Clin. Immunol. 121, 144–158 (2006). 37. Jutz, S. et al. Assessment of costimulation and coinhibition in a triple 7. Jabara, H. H. et al. A missense mutation in TFRC, encoding transferrin parameter T cell reporter line: simultaneous measurement of NF-kappaB, receptor 1, causes combined immunodeficiency. Nat. Genet. 48,74–78 (2016). NFAT and AP-1. J. Immunol. Methods 430,10–20 (2016). 8. Lelliott, P. M., McMorran, B. J., Foote, S. J. & Burgio, G. Erythrocytic iron 38. Fasano, M. et al. The extraordinary ligand binding properties of human serum deficiency enhances susceptibility to Plasmodium chabaudi infection in mice albumin. IUBMB Life 57, 787–796 (2005). carrying a missense mutation in transferrin receptor 1. Infect. Immun. 83, 39. Francis, G. L. Albumin and mammalian cell culture: implications for 4322–4334 (2015). biotechnology applications. Cytotechnology 62,1–16 (2010). 9. Conway, A. J. et al. Characterization of Tfrc-mutant mice with microcytic 40. Larsen, M. T., Kuhlmann, M., Hvam, M. L. & Howard, K. A. Albumin-based phenotypes. Blood Adv. 2, 1914–1922 (2018). drug delivery: harnessing nature to cure disease. Mol. Cell Ther. 4,1–12 (2016). 10. Beaven, G. H., Chen, S. H., d’ Albis, A. & Gratzer, W. B. A spectroscopic study 41. Nyman, D. W. et al. Phase I and pharmacokinetics trial of ABI-007, a novel of the haemin–human-serum-albumin system. Eur. J. Biochem. 41, 539–546 nanoparticle formulation of paclitaxel in patients with advanced (1974). nonhematologic malignancies. J. Clin. Oncol. 23, 7785–7793 (2005). 11. Wardell, M. et al. The atomic structure of human methemalbumin at 1.9 A. 42. Miele, E., Spinelli, G. P., Miele, E., Tomao, F. & Tomao, S. Albumin-bound Biochem. Biophys. Res. Commun. 291, 813–819 (2002). formulation of paclitaxel (Abraxane ABI-007) in the treatment of breast 12. Dutra, F. F. & Bozza, M. T. Heme on innate immunity and inflammation. cancer. Int. J. Nanomed. 4,99–105 (2009). Front. Pharm. 5, 115 (2014). 43. Morrison, K. C. & Hergenrother, P. J. Whole cell microtubule analysis by flow 13. Merlot, A. M., Kalinowski, D. S. & Richardson, D. R. Unraveling the mysteries cytometry. Anal. Biochem. 420,26–32 (2012). of serum albumin-more than just a serum protein. Front. Physiol. 5, 299 44. Lee, K. S. et al. Hemin interactions and alterations of the subcellular (2014). localization of prion protein. J. Biol. Chem. 282, 36525–36533 (2007). 14. Martins, R. et al. Heme drives -induced susceptibility to infection via 45. Monzani, E. et al. Enzymatic properties of human hemalbumin. Biochim. disruption of phagocyte functions. Nat. Immunol. 17, 1361–1372 (2016). Biophys. Acta 1547, 302–312 (2001). 15. Grinberg, L. N., O’Brien, P. J. & Hrkal, Z. The effects of heme-binding 46. Wessling-Resnick, M. Crossing the iron gate: why and how transferrin proteins on the peroxidative and catalatic activities of hemin. Free Radic. Biol. receptors mediate viral entry. Annu. Rev. Nutr. 38, 431–458 (2018). Med. 27, 214–219 (1999). 47. Chen, A. C., Donovan, A., Ned-Sykes, R. & Andrews, N. C. Noncanonical role 16. Morgan, W. T., Liem, H. H., Sutor, R. P. & Muller-Ebergard, U. Transfer of of transferrin receptor 1 is essential for intestinal homeostasis. Proc. Natl heme from heme-albumin to hemopexin. Biochim. Biophys. Acta 444, Acad. Sci. USA 112, 11714–11719 (2015). 435–445 (1976). 48. Ned, R. M., Swat, W. & Andrews, N. C. Transferrin receptor 1 is differentially 17. Pasternack, R. F. et al. Hemin binding to serum proteins and the catalysis of required in lymphocyte development. Blood 102, 3711–3718 (2003). interprotein transfer. Biochemistry 22, 1753–1758 (1983). 49. Senyilmaz, D. et al. Regulation of mitochondrial morphology and function by 18. Kamal, J. K. & Behere, D. V. Binding of heme to human serum albumin: stearoylation of TFR1. Nature 525, 124–128 (2015). steady-state fluorescence, circular dichroism and optical difference 50. Cao, H., Schroeder, B., Chen, J., Schott, M. B. & McNiven, M. A. The spectroscopic studies. Indian J. Biochem. Biophys. 42,7–12 (2005). endocytic fate of the transferrin receptor is regulated by c-Abl kinase. J. Biol. 19. Fanali, G. et al. Human serum albumin: from bench to bedside. Mol. Asp. Chem. 291, 16424–16437 (2016). Med. 33, 209–290 (2012). 51. Kamal, A. J. & Behere, D. V. Spectroscopic studies on human serum albumin 20. Vallelian, F. et al. Revisiting the putative role of heme as a trigger of and methemalbumin: optical, steady-state, and picosecond time-resolved inflammation. Pharm. Res. Perspect. 6, e00392 (2018). fluorescence studies, and kinetics of substrate oxidation by methemalbumin. 21. Miller, Y. I. & Shaklai, N. Kinetics of hemin distribution in plasma reveals its JBIC J. Biol. Inorg. Chem. 7, 273–283 (2002). role in lipoprotein oxidation. Biochim. Biophys. Acta 1454, 153–164 (1999). 52. Silva, A. M. & Hider, R. C. Influence of non-enzymatic post-translation 22. Radoshitzky, S. R. et al. Transferrin receptor 1 is a cellular receptor for New modifications on the ability of human serum albumin to bind iron: World haemorrhagic fever arenaviruses. Nature 446,92–96 (2007). implications for non-transferrin-bound iron speciation. Biochim. Biophys. 23. Coffin, J. M. Virions at the gates: receptors and the host–virus arms race. PLoS Acta 1794, 1449–1458 (2009). Biol. 11, e1001574 (2013). 53. Vlasova, I. I. Peroxidase activity of human : keeping the fire 24. Vinchi, F. et al. Hemopexin therapy reverts heme-induced proinflammatory under control. Molecules 23, 2561 (2018). phenotypic switching of macrophages in a mouse model of . 54. Tonoyan, L., Fleming, G. T. A., Mc Cay, P. H., Friel, R. & O’Flaherty, V. Blood 127, 473–486 (2016). Antibacterial potential of an antimicrobial agent inspired by peroxidase- 25. Yachie, A. et al. Oxidative stress causes enhanced endothelial cell injury in catalyzed systems. Front. Microbiol. 8, 680 (2017). human heme oxygenase-1 deficiency. J. Clin. Invest. 103, 129–135 (1999). 55. Montemiglio, L. C. et al. Cryo-EM structure of the human ferritin-transferrin 26. Prus, E. & Fibach, E. Flow cytometry measurement of the labile iron pool in receptor 1 complex. Nat. Commun. 10, 1121 (2019). human hematopoietic cells. Cytom. A 73,22–27 (2008). 56. Battin, C. et al. A human monocytic NF-κB fluorescent reporter cell line for 27. Breuer, W., Epsztejn, S. & Cabantchik, Z. I. Iron acquired from transferrin by detection of microbial contaminants in biological samples. PLoS ONE 12, K562 cells is delivered into a cytoplasmic pool of chelatable iron(II). J. Biol. e0178220 (2017). Chem. 270, 24209–24215 (1995). 57. Derdak, S. V. et al. Direct stimulation of T lymphocytes by immunosomes: 28. Epsztejn, S., Kakhlon, O., Glickstein, H., Breuer, W. & Cabantchik, I. virus-like particles decorated with T cell receptor/CD3 ligands plus Fluorescence analysis of the labile iron pool of mammalian cells. Anal. costimulatory molecules. Proc. Natl Acad. Sci. USA 103, 13144–13149 (2006). Biochem. 248,31–40 (1997). 58. Stockl, J., Majdic, O., Fischer, G., Maurer, D. & Knapp, W. Monomorphic 29. Gao, J. & Richardson, D. R. The potential of iron chelators of the pyridoxal molecules function as additional recognition structures on haptenated target cells isonicotinoyl hydrazone class as effective antiproliferative agents, IV: The for HLA-A1-restricted, hapten-specific CTL. J. Immunol. 167, 2724–2733 (2001). mechanisms involved in inhibiting cell-cycle progression. Blood 98, 842–850 59. Holter, W., Majdic, O., Stockinger, H. & Knapp, W. Analysis of T cell (2001). activation with a non-mitogenic anti CD3 antibody and the phorbol ester 30. May, P. M. & Bulman, R. A. The present status of chelating agents in TPA. Clin. Exp. Immunol. 62, 600–606 (1985). medicine. Prog. Med. Chem. 20, 225–336 (1983). 60. Holter, W., Majdic, O., Liszka, K., Stockinger, H. & Knapp, W. Kinetics of 31. Perrin, D. D. Masking and Demasking of Chemical Reactions. (Wiley-Liss, activation antigen expression by in vitro-stimulated human T lymphocytes. New York; 1970). Cell Immunol. 90, 322–330 (1985).

12 COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-01294-5 ARTICLE

61. De Sousa Linhares, A. et al. Therapeutic PD-L1 antibodies are more effective Correspondence and requests for materials should be addressed to J.S. than PD-1 antibodies in blocking PD-1/PD-L1 signaling. Sci. Rep. 9, 11472 (2019). Reprints and permission information is available at http://www.nature.com/reprints

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in Acknowledgements published maps and institutional affiliations. We thank Petra Waidhofer‐Söllner, Petra Cejka, and Claus Wenhardt for their expert technical assistance. This work was supported by the Austrian Science Fund (FWF) grant DK W1212. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, Author contributions adaptation, distribution and reproduction in any medium or format, as long as you give J.B., V.B., M.M., A.P., J.C., S.K., M.S., M.K., performed experiments and analyzed the appropriate credit to the original author(s) and the source, provide a link to the Creative data. R.C.G., G.J.Z., P.S., L.Ö., A.Y., H.C., H.S. provided essential reagents and gave Commons license, and indicate if changes were made. The images or other third party experimental advice. J.S. supervised the research and analyzed data. J.B. and J.S. wrote the material in this article are included in the article’s Creative Commons license, unless manuscript. indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory Competing interests regulation or exceeds the permitted use, you will need to obtain permission directly from The authors declare no competing interests. the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s42003- © The Author(s) 2020 020-01294-5.

COMMUNICATIONS BIOLOGY | (2020) 3:621 | https://doi.org/10.1038/s42003-020-01294-5 | www.nature.com/commsbio 13